This research was conducted to investigate in situ treatment of leachate by pilot-scale permeable reactive barrier (PRB) with vegetation. Two different types of PRB media, with and without the presence of ferric chloride sludge, for the removal of pollutants were examined. The composite media of PRB comprised a clay and sand mixture of 40:60%w/w (system 1) and a clay, ferric chloride sludge and sand mixture of 30:10:60%w/w (system 2). The system was operated at a hydraulic loading rate of 0.028 m3/m2.d and hydraulic retention time of 10 days. The results showed that the performance of system 2 was better in terms of pollutant removal efficiencies, with average biochemical oxygen demand, chemical oxygen demand and total Kjeldahl nitrogen removals of 76.1%, 68.5% and 73.5%, respectively. Fluorescence excitation-emission matrix analyses of water samples and sequential extraction of PRB media suggested the removal of humic substances through the formation of iron–organic complex. Greenhouse gas (GHG) emissions during the treatment of PRB were 8.2–52.1 mgCH4/m2.d, 69.1–601.8 mgCO2/m2.d and 0.04–0.99 mgN2O/m2.d. The use of system 2 with vegetation resulted in lower GHG emissions. The results show that PRB with vegetation could be used as a primary treatment for leachate from closed landfill sites.

INTRODUCTION

Leachate is a highly concentrated wastewater generated as a by-product from municipal solid waste landfill. It poses a threat to the environment in terms of air, surface water and groundwater pollution. Conventional leachate treatment systems available in developing countries are generally low-cost technologies such as stabilization ponds or aerated lagoons (Johannessen & Boyer 1999). Nevertheless, those treatment technologies are usually implemented only during the operation stage of solid waste disposal sites. Most of the treatment systems are not well operated, or are poorly maintained, due to lack of financial support for care and maintenance in landfill sites after their closure. The control of leachate migration off the site after landfill closure is not an easy task because of high variations in leachate quantity and characteristics with time. In order to treat landfill leachate efficiently, treatment systems should be capable of removing easily biodegradable – as well as recalcitrant – organic compounds present in leachate. They also need to cope with the high nitrogen content in leachate. In this study, we develop a passive control system for closed landfill sites. The concept of the treatment technology is based on a combination of constructed wetlands, which have been using polishing treatments for various wastewater types, including landfill leachate (Sim et al. 2013) and permeable reactive barriers (PRB), for remediation of contaminated groundwater (Hou et al. 2014). A subsurface flow constructed wetland that utilizes coarse media with high permeability and vegetation has been proved to be an efficient treatment method for fresh landfill leachate (Chiemchaisri et al. 2009). Its main treatment functions include biological treatment using attached growth on the media and plant root, and plant uptake of nutrients. Nevertheless, the performance of the constructed wetland was limited when being applied with stabilized leachate containing recalcitrant organic compounds and a high ammonia concentration. PRB technology using low permeable media like clay and sand mixture has also been successfully applied for reducing groundwater pollution from solid waste disposal sites (Dong et al. 2009). PRB utilizes reactive media to prevent contaminated water pollutants by physical, chemical and/or biological processes. The treatment processes can include precipitation, sorption, oxidation/reduction, fixation or degradation to remove the contaminants from the water (McMahon et al. 1999). An improved PRB performance was attributed to chemical reactions between target pollutants and reactive media while the biodegradation process proceeded simultaneously. Applications of PRB for removing chemical oxygen demand (COD) (Chung et al. 2007; Liu et al. 2011), ammonia and adsorbable organic halogens (Van Nooten et al. 2008), heavy metals (Bartzas et al. 2006; Chung et al. 2007; Komnitsas et al. 2013) from landfill leachate have been investigated. The types of reactive media used include zero-valent iron (ZVI) (Bartzas et al. 2006), ZVI with fly ash/red mud (Komnitsas et al. 2013), ZVI with activated carbon and modified bentonite (Liu et al. 2011), clinoptilolite and activated carbon (Van Nooten et al. 2008). With a different approach, this research aims at utilizing available materials (native clay) and wastes (ferric chloride sludge) at a solid waste disposal site for the development of PRB with native vegetation as an in situ pollution control technology for closed landfills. The main purpose for the introduction of ferric chloride sludge in reactive media in this study was to investigate its effect on the suppression of methane production during treatment (Sylvia & Schnell 2000). This technology can be practically implemented by constructing a shallow PRB surrounding the waste disposal area as a primary treatment so that infiltrated leachate can be preliminarily treated before being sent to the main leachate treatment system. The technology would be suitable for a landfill that is placed above a low permeable soil layer at which, horizontal movement or overflow of leachate needs to be controlled. It is also anticipated that vegetation could enhance the performance of PRB by improving soil porosity and oxygen diffusion for microbial activities. In our previous investigations, local vegetation has provided a positive effect on leachate treatment in soil-plant system, while also helping to reduce greenhouse gas (GHG) emissions during treatment (Suwunpukdee et al. 2013).

MATERIALS AND METHODS

Experimental system

The experimental system is located at a municipal solid waste disposal site in Thailand. The site receives more than 900 tons of waste daily and has been in operation for more than 30 years (from 1982 to present). Two pilot-scale PRB units were constructed at the periphery of a landfill area of 10.9 ha, which has been closed since 2005. The ages of disposed solid waste in the landfill ranged from 10 to 30 years. The PRB units were applied to the treatment of real landfill leachate pumped from a leachate collection pit, which receives drained leachate from the closed landfill area. The dimension of each experimental unit is 1 m width, 2 m length, and 1 m depth, as the details show in Figure 1. The inlet and outlet zones of the experimental unit were filled with 30–60 mm gravel of 0.80 m depth. In between the inlet and outlet zones, PRB media of 1.5 m length and 0.8 m depth was installed. Vegetation was grown on the surface of the PRB media. Two PVC pipes (25 mm inch diameter) were provided for sampling water at every 0.5 m interval along the treatment pathway under the vegetation root zone.

Figure 1

Schematic of pilot-scale permeable reactive barrier (PRB) with vegetation.

Figure 1

Schematic of pilot-scale permeable reactive barrier (PRB) with vegetation.

Guinea grass (Panicum maximum TD 58) was used as vegetation in both units with an initial plant density of 20 rhizomes/m2. The plants were obtained locally at the same solid waste disposal site as they were found to be effective for landfill leachate treatment (Suwunpukdee et al. 2013).

PRB media

Clay and sand mixture was used as base material in PRB media at 40:60 (%w/w) in system 1. In order to evaluate the effect of ferric chloride sludge, clay: ferric chloride sludge: sand mixture at 30:10:60 (%w/w) was provided in system 2. The properties of the media used in this study are shown in Table 1. PRB media made of clay was obtained from local soil, and ferric chloride sludge was brought from the sludge storage pond of a leachate treatment plant located in the same solid waste disposal site. The sludge was produced during chemical coagulation of leachate collected from the closed landfill area by adding ferric chloride at a dose of 1.5–2.0 g/l. The produced sludge that settled in the clarifier was drained to the sludge storage pond for disposal. More details for performance of the leachate treatment system are reported in Theepharaksapan et al. (2011). The sludge obtained from the sludge pond was air-dried prior to its use in the PRB media. The air-dried ferric chloride sludge had a bulk density of 1,039 kg/m3 with porosity of 0.61. It contained total organic carbon (TOC) and total nitrogen (TN) of 17.3 g/kg and 2,780 mg/kg, respectively. The extractable iron in the ferric chloride sludge was 3.54 g/kg sludge in which it was mainly in oxide (68.5%) and organic bound (20.2%) forms.

Table 1

Properties of media

Parameter (unit) PRB 1 (clay:sand) PRB 2 (clay:ferric chloride sludge:sand) 
pH (−) 6.45 6.84 
Porosity (%) 40.16 40.14 
Salinity, NaCl (%) 0.27 0.49 
Extractable Fe (g/kg media) 2.80 6.30 
Electrical conductivity (dS/m) 2.49 3.87 
Bulk density (kg/m31,620 1,540 
Hydraulic conductivity, k (m/s) 2.93 × 10−6 5.72 × 10−6 
Soil texture Sandy loam Sandy loam 
 Sand (%) 65.36% 69.36% 
 Silt (%) 11.55% 5.22% 
 Clay (%) 23.09% 25.42% 
Parameter (unit) PRB 1 (clay:sand) PRB 2 (clay:ferric chloride sludge:sand) 
pH (−) 6.45 6.84 
Porosity (%) 40.16 40.14 
Salinity, NaCl (%) 0.27 0.49 
Extractable Fe (g/kg media) 2.80 6.30 
Electrical conductivity (dS/m) 2.49 3.87 
Bulk density (kg/m31,620 1,540 
Hydraulic conductivity, k (m/s) 2.93 × 10−6 5.72 × 10−6 
Soil texture Sandy loam Sandy loam 
 Sand (%) 65.36% 69.36% 
 Silt (%) 11.55% 5.22% 
 Clay (%) 23.09% 25.42% 

Leachate

Leachate was fed into the system at a hydraulic loading rate (HLR) of 0.028 m3/m2.d, equivalent to a hydraulic retention time (HRT) of 10 d. This experimental condition was pre-determined to be sufficient for leachate purification in our previous study (Chiemchaisri et al. 2009). The performance of PRB in leachate purification was monitored over 240 days by comparing influent, treated water along the treatment pathway, and effluent qualities on weekly basis. The analyses of water quality parameters were performed according to Standard Methods for the Examination of Water and Wastewater (APHA 2005). The heavy metal (Cr, Cu, Ni, Pb, Cd) concentrations found in leachate were detected at much lower concentrations when compared to the effluent standard of Thailand and therefore they are not focused on in this study. The water quality data were analysed by Student's t-test and one-way analysis of variance (ANOVA) to compare the statistical difference between two systems at a 95% confidence level. The characteristics of leachate used in this study are shown in Table 2. The leachate exhibits the characteristics of stabilized leachate with an alkaline nature (pH > 8) and low biochemical oxygen demand (BOD)/COD of 0.34. Other major inorganic constituents are NH3 and total dissolved solids (TDS) while having low NO2, NO3 and total phosphorus (TP) concentrations.

Table 2

Characteristics of leachate used in this study

Parameter Range Mean 
pH 7.5–8.8 8.3 
EC (dS/m) 11.5–25.9 19.2 
NaCl (%) 0.19–0.50 0.35 
BOD (mg/l) 704–1,434 922 
COD (mg/l) 2,200–3,145 2,694 
SS (mg/l) 230–810 575 
TDS (mg/l) 8,140–13,520 10,610 
NH (mg as N/l) 56.7–179 102 
TKN (mg as N/l) 112–296 191 
NO2 (mg as N/l) 0.1–1.3 0.4 
NO3 (mg as N/l) 0.2–3.5 1.2 
TP (mg/l) 4.8–11.6 5.2 
Parameter Range Mean 
pH 7.5–8.8 8.3 
EC (dS/m) 11.5–25.9 19.2 
NaCl (%) 0.19–0.50 0.35 
BOD (mg/l) 704–1,434 922 
COD (mg/l) 2,200–3,145 2,694 
SS (mg/l) 230–810 575 
TDS (mg/l) 8,140–13,520 10,610 
NH (mg as N/l) 56.7–179 102 
TKN (mg as N/l) 112–296 191 
NO2 (mg as N/l) 0.1–1.3 0.4 
NO3 (mg as N/l) 0.2–3.5 1.2 
TP (mg/l) 4.8–11.6 5.2 

Organic matter characterization

The fluorescence excitation-emission matrix (FEEM) spectroscopy technique (Jasco FP-8200 spectrofluorometer, Tokyo, Japan) was used for the characterization of organic matter in influent and effluent samples following the methodology described in Sanguanpak et al. (2013). All samples were diluted to a dissolved organic carbon (DOC) concentration of 10 mg/l and adjusted to pH 7 before analysis. FEEM spectra were collected at excitation (Ex) wavelengths from 200 to 500 nm and emission (Em) wavelengths from 250 to 600 nm with 5 nm increments. The spectra were scanned with a 5 nm slit bandwidth at a scan rate of 2,000 nm/min. The spectrum of de-ionized (DI) water was recorded as blank, and the equipment was auto-calibrated before analysis.

Analyses of PRB media and vegetation

PRB media were collected as composite samples between inlet, middle and outlet parts and characterized for pH, electrical conductivity (EC), organic matter, organic carbon, moisture content, NH3, NO3, total Kjeldahl nitrogen (TKN) and TP at the beginning and the end of experiment. Water permeability of media was determined in the experiment unit before and after the experiment by using the constant head method. The growth of plants was determined in terms of shoot height, root length, number of leaves and total dry weight at the beginning and end of the experiment.

The iron forms in sludge and used media after the operation were characterized by the sequential extraction method (Konradi et al. 2005). For water soluble and exchangeable forms, 1 g of sample was extracted with 20 ml DI water and 1.0 M MgCl2 solution for 1 h, respectively. For the carbonate bound form, the carbonates in the residue from the previous step were extracted with 20 ml of a 1.0 M NaAc solution adjusted to a pH of 5.0 with HAc by continuously shaking for 4 h. To determine the oxide form, the residue of the carbonate extraction step was extracted by shaking with 50 ml of a 0.04 M NH2OH·HCl/25% HAc solution for 5.5 h at 96 °C. Then, the organic bound form was determined by pouring 7.5 ml of a 0.02 M HNO3 solution and 12.5 ml of a 30% H2O2 solution adjusted to a pH of 2.0, and then providing continuous agitation for 2 h at 85 °C. An additional volume of 7.5 ml of the 30% H2O2 solution adjusted to a pH of 2.0 was then added, while maintaining continuous agitation and a temperature of 85 °C for another 3 h. This solution was then cooled to room temperature. An aliquot of 12.5 ml of a 3.2 M NH4Ac solution was added and shaken for 30 min. The residue fraction was determined by using residue from organic extraction transferred to a digestion vessel, which was digested with 10 ml each of concentrated HNO3, and HF at 220 °C, to a moist bead. The moist bead was taken up in 25 ml 4 M HC1 at 100 °C for 30 minutes. The final volume was adjusted to 25 ml with DI water. The samples at each extraction step were subjected to analysis of iron concentrations by atomic adsorption spectroscopy.

Greenhouse gas emissions

During the treatment, GHG emissions were also evaluated using the closed flux chamber technique on a monthly basis. The flux measurements were performed using a chamber of 0.3 m diameter and a 0.5 m height cover over PRB media. In the case of measurement over the area with vegetation, a chamber with 1.0 m height was used. The determination of gas emissions was conducted over bare soil (non-rhizospheric zone) and soil with plants (rhizospheric zone) at a distance of 0.25, 0.75 and 1.25 m length: the so-called inlet, middle and outlet point of soil media, respectively, for comparison to investigate the effect of plants on reducing GHG emissions. Methane and carbon dioxide in the gas samples collected at 15-minute intervals from the chamber were analysed by Shimadzu GC-14B (Kyoto, Japan), whereas nitrous oxide was analysed by Shimadzu GC-Clarus 580. The gas flux was determined from concentration increase rates in the chamber as described in the following equation: 
formula
1
where F is the gas flux (mg/m2.d) at 25 °C, V is the chamber volume (m3), A is the area enclosed by the chamber (m2), ΔCt is the gas concentration gradient (mg/m3.d) and T is the temperature of the air within the chamber (°C).

RESULTS AND DISCUSSION

PRB treatment performance

Figure 2 shows the variation of BOD, COD and TKN concentrations and their removal during the experimental period of 240 days. The volume of leachate fed to each experimental unit was 42 l/d. After treatment, the effluent flow from system 1 and system 2 had an average flow of 31 and 24 l/d, resulting in a 26% and 43% reduction in leachate volume, respectively. The leachate volume was mainly reduced in system 1 by evaporation, whereas system 2 provided further reduction through plant transpiration. During the operation, it was found that the organic and nitrogen concentrations in the PRB effluent were kept relatively stable in both systems while their concentrations in the feeding leachate fluctuated. Table 3 shows average effluent qualities from the pilot-scale PRB over 240 days. In terms of treatment performance, moderate organic and nutrient (N, P) removals of about 50–80% were achieved in the PRB system. It was found that the removal rates of pollutants were different between the experimental units using different media. For most pollutants, except color, the removals were higher in the experimental unit in which iron sludge was added into the clay:sand mixture. On average, BOD, COD and TKN removals were 76.1%, 68.5% and 73.5%, respectively; meanwhile, those for the clay:sand mixture alone were 69.5, 63.3 and 66.9%, respectively.

Table 3

Effluent characteristics and steady state removal efficiencies

  PRB 1 (clay:sand)
 
PRB 2 (clay:iron sludge:sand)
 
  
Parameter Port 1 Port 2 Eff. Port 1 Port 2 Eff. % removal PRB1 % removal PRB2 
pH 7.86 7.76 7.71a 8.22 8.07 7.79a – – 
EC (dS/m) 13.8 11.5 10.9a 13.1 10.9 9.9a 43.3 48.4 
Color (unit) 572 407 282a 622 467 349a 70.1 61.5 
NaCl (%) 0.26 0.21 0.18a 0.24 0.20 0.16a 48.3 54.1 
ORP (mV) − 231 − 205 − 159b − 230 − 184 − 155b – – 
BOD (mg/l) 577 399 281a 544 342 221a 69.5 76.1 
COD (mg/l) 1,824 1,331 989a 1,630 1,196 849a 63.3 68.5 
TOC (mg/l) – – 561a – – 490a 43.4 50.4 
SS (mg/l) – – 150a – – 114a 73.9 80.2 
TDS (mg/l) – – 5,475a – – 4,887a 48.4 53.9 
NH3-N (mg/l) 55 52 31b 50 45 28b 69.5 73.4 
TKN (mg/l) 147 112 63a 134 92 51a 66.9 73.5 
NO2-N(mg/l) 0.2 0.15 0.05b 0.2 0.16 0.04b 76.2 80 
NO3-N(mg/l) 0.9 0.5 0.24b 0.86 0.66 0.21b 78.8 84 
TP (mg/l) 3.30 3.03 1.44b 3.51 3.17 1.36b 71.7 74.4 
  PRB 1 (clay:sand)
 
PRB 2 (clay:iron sludge:sand)
 
  
Parameter Port 1 Port 2 Eff. Port 1 Port 2 Eff. % removal PRB1 % removal PRB2 
pH 7.86 7.76 7.71a 8.22 8.07 7.79a – – 
EC (dS/m) 13.8 11.5 10.9a 13.1 10.9 9.9a 43.3 48.4 
Color (unit) 572 407 282a 622 467 349a 70.1 61.5 
NaCl (%) 0.26 0.21 0.18a 0.24 0.20 0.16a 48.3 54.1 
ORP (mV) − 231 − 205 − 159b − 230 − 184 − 155b – – 
BOD (mg/l) 577 399 281a 544 342 221a 69.5 76.1 
COD (mg/l) 1,824 1,331 989a 1,630 1,196 849a 63.3 68.5 
TOC (mg/l) – – 561a – – 490a 43.4 50.4 
SS (mg/l) – – 150a – – 114a 73.9 80.2 
TDS (mg/l) – – 5,475a – – 4,887a 48.4 53.9 
NH3-N (mg/l) 55 52 31b 50 45 28b 69.5 73.4 
TKN (mg/l) 147 112 63a 134 92 51a 66.9 73.5 
NO2-N(mg/l) 0.2 0.15 0.05b 0.2 0.16 0.04b 76.2 80 
NO3-N(mg/l) 0.9 0.5 0.24b 0.86 0.66 0.21b 78.8 84 
TP (mg/l) 3.30 3.03 1.44b 3.51 3.17 1.36b 71.7 74.4 

aDifference between system 1 and system 2 with statistical significance (p < 0.05).

bNo difference between system 1 and system 2 with statistical significance (p > 0.05).

Figure 2

Variation in BOD, COD and TKN concentrations, and their removal efficiencies.

Figure 2

Variation in BOD, COD and TKN concentrations, and their removal efficiencies.

Moderate dissolved solids removals in terms of EC (48.4%) and salinity (54.1%) were also observed in the PRB system, suggesting the removal of salts through precipitation either in media or the plant root zone. Moreover, vegetation could also help to improve organic and TKN removal, possibly through the enhancement of oxygen transfer into the soil by the plant's root system, which subsequently promoted aerobic biodegradation by soil microorganisms. Considering the effluent qualities from PRB treatment, polishing treatment of PRB effluent would be required in order to be discharged safely into the natural water environment. It is suggested that this PRB technology can serve as a primary treatment to reduce the volume of leachate and pollutant concentrations prior to its treatment in the main leachate treatment system. It is also noted that pollutant removal in the PRB with ferric chloride sludge was not much improved over the system using base materials because the amount of ferric chloride sludge added to PRB media was limited to only 10% w/w as the media was designed based on pre-determined hydraulic permeability. Further investigation on the effect of the sludge amount would help in determining an appropriate media mixture ratio to achieve higher pollutant removals.

Pollutant transformation in PRB

Figure 3 shows the FEEM spectrum of influent and effluent from PRB systems, suggesting that the majority of organic matter presented in leachate was a humic-like substance with an observed peak at ex245–260/em408–460 nm. The relative intensity of the FEEM peak suggested that PRB 2 had higher removal efficiencies of the humic-like substance (77%) when compared to PRB 1 (62%). The enhanced treatment in ferric chloride sludge amended media was possibly due to organic adsorption onto sludge particles as suggested by the predominance of iron–organic complex from the sequential extraction of iron in the reactive media (Table 4).

Table 4

Amount of extracted iron (Fe) from PRB media in different forms

  Fe (g/kg media)
 
System/zone II III IV VI 
PRB 1- inlet 0.006 1.425 0.034 0.380 1.076 0.011 
- middle 0.013 1.575 0.022 0.271 1.164 0.010 
- outlet 0.016 2.100 0.023 0.266 0.883 0.068 
PRB 2- inlet 0.055 0.575 1.669 2.598 0.896 0.608 
- middle 0.006 3.175 1.520 2.555 3.836 0.273 
- outlet 0.020 2.275 0.262 1.193 2.364 1.484 
  Fe (g/kg media)
 
System/zone II III IV VI 
PRB 1- inlet 0.006 1.425 0.034 0.380 1.076 0.011 
- middle 0.013 1.575 0.022 0.271 1.164 0.010 
- outlet 0.016 2.100 0.023 0.266 0.883 0.068 
PRB 2- inlet 0.055 0.575 1.669 2.598 0.896 0.608 
- middle 0.006 3.175 1.520 2.555 3.836 0.273 
- outlet 0.020 2.275 0.262 1.193 2.364 1.484 

I = water soluble form, II = exchangeable form, III = carbonate bound form, IV = iron and manganese bound form, V = organic bound form, VI = residue.

Figure 3

FEEM spectrum of influent and effluent from PRB treatment. (a) Influent, (b) effluent-PRB 1, (c) effluent-PRB 2. Y-axis shows relative intensity (a.u.) of FEEM spectrum.

Figure 3

FEEM spectrum of influent and effluent from PRB treatment. (a) Influent, (b) effluent-PRB 1, (c) effluent-PRB 2. Y-axis shows relative intensity (a.u.) of FEEM spectrum.

The results of chemical analyses of media before and after the experiment are shown in Table 5. It was found that carbon and nitrogen were partially accumulated in the media, resulting in an increase in TOC and TN in both systems. It is anticipated that organic matter is first adsorbed onto the soil and then subsequently biodegraded by microorganisms. In comparison, PRB 2 had lower initial organic carbon than system 1 due to the presence of iron sludge. After the experiment, the accumulation of pollutants in the PRB 2 matrix were found to be higher than PRB 1 in terms of organic carbon, whereas nitrogen accumulations in both systems were not much different. A significant increase in salt content in the media was also observed, possibly resulting from salt crystallization in the media when part of the water evaporated during the treatment. Moreover, plants can uptake nutrients, mainly nitrogen, for their growth. Accumulation of pollutants and biomass in the media led to a decrease in their hydraulic permeability.

Table 5

Characteristics of media before and after the experiment

  Before
 
After
 
Parameter PRB 1 PRB 2 PRB 1 PRB 2 
pH (–) 6.45 6.84 6.5 7.1 
NaCl (%) 0.27 0.24 0.49 0.78 
EC (dS/m) 2.49 3.87 13.9 11.8 
TN (mg/kg) 188.3 417.2 903.3 1,166.6 
TOC (g/kg) 13.64 11.96 15.36 14.54 
Permeability, k (m/s) 2.93 × 10−6 5.72 × 10−6 2.26 × 10−7 1.73 × 10−7 
  Before
 
After
 
Parameter PRB 1 PRB 2 PRB 1 PRB 2 
pH (–) 6.45 6.84 6.5 7.1 
NaCl (%) 0.27 0.24 0.49 0.78 
EC (dS/m) 2.49 3.87 13.9 11.8 
TN (mg/kg) 188.3 417.2 903.3 1,166.6 
TOC (g/kg) 13.64 11.96 15.36 14.54 
Permeability, k (m/s) 2.93 × 10−6 5.72 × 10−6 2.26 × 10−7 1.73 × 10−7 

Plant growth

During the operation of the PRB system, the growth of the plant (Panicum maximum TD 58) was also studied. During the whole experimental period, the vegetation was allowed to grow without being harvested. It was found that the growth of plants in system 2 was better than system 1 as shown in terms of plant height, root length, number of leaves and plant dry weight (Table 6). From these results, it can be concluded that the provision of iron sludge in PRB media did not have a negative impact on plant growth, but on the other hand, it slightly promoted plant growth. The promotion of plant growth could be due to an improvement in the physical properties of the media for plant growth, or, the addition of ferric iron may also help in promoting the photosynthesis mechanism of plant chlorophyll. The higher growth of plants in PRB 2 also promoted water removal through evapotranspiration, resulting in lower effluent volume obtained from PRB 2 compared to PRB 1. At the end of the experiment, adsorption of heavy metals (Cr, Cu, Ni, Pb, Cd) in plants was found to be insignificant, and the harvested plants can be safely disposed of in landfill.

Table 6

Growth of Panicum maximum

  After
 
Parameter Before PRB 1 PRB 2 
Height (m) 0.15 1.01 1.38 
Root (m) 0.12 0.21 0.34 
No. of leaves 55 84 
Dry weight (g) 33 2,441 3,254 
  After
 
Parameter Before PRB 1 PRB 2 
Height (m) 0.15 1.01 1.38 
Root (m) 0.12 0.21 0.34 
No. of leaves 55 84 
Dry weight (g) 33 2,441 3,254 

Greenhouse gas emissions

Table 7 compares average methane, carbon dioxide and nitrous oxide emission rates from soil (non-rhizospheric zone) and soil-plant (rhizospheric zone) in PRB with different media. Methane, carbon dioxide and nitrous oxide emissions during the treatment of PRB were 8.0–52.6, 69.1–570.2 and 0.04–0.69 mg/m2.d, respectively. Among them, carbon dioxide emission was highest, followed by methane and nitrous oxide. For all gases, the emission rates from bare soil were higher than the soil-plant location, indicating that the plant helped in reducing GHG emissions, possibly due to beneficial effects from oxygen transfer out of their root system. Emissions were also found to be highest near the inlet part and gradually decreased along the treatment pathway. Comparing both systems, system 2 had lower emission rates than system 1 for all gases. These results suggested that an introduction of iron sludge into the media helped mitigate GHG emissions during treatment. This is an expected mechanism as ferric chloride amended sludge media could promote the activity of Fe3+ reducing bacteria, resulting in a switch in electron flow from methanogenesis to Fe3+ reduction (Frenzel et al. 1999). Furthermore, plant–microbe interactions could lead to an efficient iron oxidation and reduction reaction in an oxidized rhizosphere environment (Roden & Wetzel 1996).

Table 7

Emission rate (mg/m2.d) of greenhouse gases from the PRB system

  Methane
 
Carbon dioxide
 
Nitrous oxide
 
System Location Range Average Range Average Range Average 
Bare soil 16.3–52.6 31.2 215.7–570.2 380.7 0.21–0.69 0.42 
Soil-plant 14.2–46.8 26.6 155.0–536.0 316.1 0.14–0.53 0.31 
Bare soil 11.3–41.7 22.6 112.5–508.9 293.6 0.07–0.43 0.21 
Soil-plant 8.0–37.0 17.8 69.1–373.3 259.6 0.04–0.41 0.19 
  Methane
 
Carbon dioxide
 
Nitrous oxide
 
System Location Range Average Range Average Range Average 
Bare soil 16.3–52.6 31.2 215.7–570.2 380.7 0.21–0.69 0.42 
Soil-plant 14.2–46.8 26.6 155.0–536.0 316.1 0.14–0.53 0.31 
Bare soil 11.3–41.7 22.6 112.5–508.9 293.6 0.07–0.43 0.21 
Soil-plant 8.0–37.0 17.8 69.1–373.3 259.6 0.04–0.41 0.19 

CONCLUSIONS

PRB with clay and sand as base materials was applied to the treatment of leachate at a closed landfill site in Thailand. When the system was operated at a hydraulic loading rate of 0.028 m3/m2.d or hydraulic retention time of 10 days, BOD, COD and TKN removals in PRB media containing clay, ferric chloride sludge and a sand mixture of 30:10:60%w/w were 76.1, 68.5 and 73.5%, respectively. More than a 40% reduction in leachate volume took place in the PRB with vegetation. Improvement of humic substances through the formation of iron–organic complex in the PRB media was observed. Introduction of vegetation (Panicum maximum TD 58) helped reduce GHG emissions during the PRB treatment by enhancing oxygen transfer into the system. This PRB technology can serve as primary treatment to reduce the volume of leachate and pollutant concentrations prior to its treatment in the main leachate treatment system.

ACKNOWLEDGEMENT

This research work was financially supported by the National Research Council of Thailand (NRCT) under the Asian Core Program.

REFERENCES

REFERENCES
APHA
2005
Standard Methods for the Examination of Water and Wastewater
20th edn.
American Public Health Association/American Water Works Association/Water Environment Federation
,
Washington DC, USA
.
Bartzas
G.
Komnitsas
K.
Paspaliaris
L.
2006
Laboratory evaluation of FeO barriers to treat acidic leachates
.
Minerals Engineering
19
(
5
),
505
514
.
Chiemchaisri
C.
Chiemchaisri
W.
Junsod
J.
Threedeach
S.
Wicramarachchi
P. N.
2009
Greenhouse gas emission from constructed wetland for treating landfill leachate in the tropics
.
Bioresource Technology
100
,
3808
3814
.
Chung
H. L.
Kim
S. K.
Lee
Y. S.
Yu
J.
2007
Permeable reactive barrier using atomized slag material for treatment of contaminants from landfills
.
Groundwater Journal
11
(
2
),
137
145
.
Hou
G.
Liu
F.
Liu
M.
Kong
X.
Li
S.
Chen
L.
Colberg
P. J. S.
Jin
S.
Chen
S.
2014
Performance of a permeable reactive barrier for in-situ removal of ammonia in groundwater
.
Water Science and Technology: Water Supply
14
(
4
),
585
592
.
Johannessen
L. M.
Boyer
G.
1999
Observations of Solid Waste Landfills in Developing Countries: Africa, Asia and Latin America, Urban and Local Government Working Paper
.
Urban Development Division, The World Bank
.
Komnitsas
K.
Bazdanis
G.
Sahinkaya
E.
Zaharaki
D.
2013
Removal of heavy metals from leachates using organic/inorganic permeable reactive barriers
.
Desalination & Water Treatment
51
(
13–15
),
3052
3059
.
Konradi
E. A.
Frentiu
T.
Michaela
P.
Cordos
E.
2005
Use of sequential extraction to assess metal fractionation in soils from Bozanta Mare, Romania
.
Acta Universitatis Cibiniensis Seria F Chemia
8
,
5
12
.
Liu
J. J.
Zeng
N. S.
Xu
W. X.
2011
Effects of the use of permeable barrier for landfill leachate treatment
.
Journal of Water and Environment Technology
9
(
2
),
209
214
.
Sim
C. H.
Quek
B. S.
Shutes
R. B. E.
Goh
K. H.
2013
Management and treatment of landfill leachate by a system of constructed wetlands and ponds in Singapore
.
Water Science and Technology
68
(
5
),
1114
1122
.
Suwunpukdee
C.
Chiemchaisri
C.
Chiemchaisri
W.
Tudsri
S.
2013
Utilization of concentrated leachate for plant cultivation on municipal solid waste landfill
. In:
Southeast Asian Water Environment 5
,
IWA Publishing
,
UK
.
Sylvia
U.
Schnell
S.
2000
Suppression of methane emission from rice paddies by ferric iron fertilization
.
Journal of Soil Biology & Biochemistry
32
,
1811
1814
.